nanoscale zerovalent iron for source remediation

8/6/2019 Nanoscale Zerovalent Iron for Source Remediation

1/54

Printed on recycled paper

CONTRACTREPORT

CR-05-007-ENV

COST AND PERFORMANCE REPORT

NANOSCALE ZERO-VALENT IRON TECHNOLOGIES FOR

SOURCE REMEDIATION

by

Arun GavaskarLauren TatarWendy Condit

September 2005

Approved for public release; distribution is unlimited.

ENGINEERING SERVICE CENTERPort Hueneme, California 93043-4370


2/54

REPORT DOCUMENTATION PAGEForm Approved

OMB No. 0704-0811

The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources,gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of thiscollection of information, including suggestions for reducing the burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations andReports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no personshall be subject to any penalty for failing to comply with a collection of information, it if does not display a currently valid OMB control number.

PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.

1. REPORT DATE (DD-MM-YYYY) 2. REPORT TYPE 3. DATES COVERED (From To)

September 2005 Final

4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

COST AND PERFORMANCE REPORT FOR NANOSCALE

ZER0-VALENT IRON TECHNOLOGIES FOR COURCE

REMEDIATION

6. AUTHOR(S) 5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

Arun Gavaskar, Lauren Tatar, and Wendy Condit

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESSES 8. PERFORMING ORGANIZATION REPORT NUMBER

Commanding Officer

Naval Facilities Engineering Service Center

1100 23rd AvePort Hueneme, CA 93043-

CR-05-007-ENV

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITORS ACRONYM(S)

11. SPONSOR/MONITORS REPORT NUMBER(S)

12. DISTRIBUTION/AVAILABILITY STATEMENT

Approved for public release; distribution is unlimited.

13. SUPPLEMENTARY NOTES

14. ABSTRACT

This cost and performance report is a compilation of technical and performance data from three recent Navydemonstration projects involving the use of microscale or nanoscale zero-valent iron (NZVI) for treatment of dense,

nonaqueous-phase liquid (DNAPL) source zones.

15. SUBJECT TERMS

Nano-scale zero-valent (NZVI), iron, groundwater, remediation

16. SECURITY CLASSIFICATION OF: 19a. NAME OF RESPONSIBLE PERSON

a. REPORT b. ABSTRACT c. THIS PAGE

17. LIMITATION OF

ABSTRACT

18. NUMBER OFPAGES

19b. TELEPHONE NUMBER (include area code)U U U U 54

Standard Form 298 (Rev. 8/98)Prescribed by ANSI Std. Z39.18


3/54

FINAL

COST AND PERFORMANCE REPORT

NANOSCALE ZERO-VALENT IRON TECHNOLOGIESFOR SOURCE REMEDIATION

Contract Number: N47408-01-D-8207Task Order: 0076

Prepared for:

Naval Facilities Engineering Service Center1100 23rd Avenue

Port Hueneme, California 93043-4301

Prepared by:

Arun Gavaskar, Lauren Tatar, and Wendy Condit

BATTELLE505 King Avenue

Columbus, Ohio 43201

August 29, 2005


4/54

EXECUTIVE SUMMARY

Zero-valent iron (ZVI) has been used in permeable reactive barriers for groundwater treat-ment for over ten years now. ZVI or elemental iron (Fe0) is a strong reducing agent that is capable ofabiotically dehalogenating several common chlorinated solvents (e.g., trichloroethene [TCE]), which arecommon pollutants at military and industrial sites. The granular ZVI used in permeable barrier applica-

tions typically consists of iron particles in the size range of

8+50 mesh, which makes the ZVI barriermore permeable than the surrounding aquifer. An emerging technology based on ZVI is the use of nano-scale zero-valent iron (NZVI) for source zone, rather than plume, treatment. In source areas, solvent maybe present as dense nonaqueous-phase liquid (DNAPL) or free-phase solvent and dissolved-phase concen-trations in the region tend to be much higher than in the downgradient plume. The finer NZVI particlesare much more reactive than granular ZVI and have the potential to quickly treat the higher concentra-tions of chlorinated volatile organic compounds (CVOC) present in source zones. Also, finer particles areeasier to inject in the soil pores than coarse particles, so the smaller particle size of NZVI helps itsdelivery.

The Navy has recently conducted NZVI field demonstrations at three sites: Hunters PointShipyard (Hunters Point), Naval Air Station (NAS) Jacksonville, and Naval Air Engineering Station

(NAES) Lakehurst. This cost and performance (C&P) report is the result of a comparative evaluation ofthe performance of NZVI injection at these three sites. At Hunters Point the iron injected was not trueNZVI, but micron-sized ZVI powder called Ferox. This powder, provided by ARS Technologies, Inc.,is expected to be less reactive compared to NZVI, but is lower in price as well, probably because of itscoarser particle size and differences in manufacture. At the Jacksonville and Lakehurst sites, a variety ofNZVI called bimetallic nanoscale particles (BNP) provided by PARS Environmental Inc. was used.Addition of trace quantities of a second metal, such as palladium, improves the reactivity of the iron stillfurther, as the second metal catalyzes the dehalogenation reactions.

The Navy conducted considerable performance monitoring at the three sites and the keyresults are summarized in this report. In general, the following are the key observations made in thiscomparative study:

At Hunters Point, two ZVI injection studies were conducted, one in the sourcearea and the other in the plume. In the first study, 16,000 lb of micron-sized ZVIpowder was made into a 265 g/L iron slurry in tap water and was injected into theDNAPL source zone by pneumatic fracturing, using nitrogen as the carrier gas.The iron-to-soil ratio achieved in the target treatment zone was 0.004. Afterinjection, ORP of the groundwater dropped to below 500 mV and pH roseabove 8, indicating that strongly reducing conditions suitable for abioticdehalogenation of TCE were generated. There was no significant formation ofcis-1,2-DCE immediately after injection, thus indicating that microbially drivenanaerobic reduction was not the primary mechanism for TCE mass removal.Longer-term monitoring over one year after injection showed some signs of a

rebound (increase) in ORP and DCE in some wells, thus indicating that the ZVIwas losing some reactivity. However, TCE levels continued to remain low andthe DCE rebound subsided eventually, thus indicating that ORP remainedreducing enough to promote biodegradation and hydrogenolysis of residuals.

In the second injection study at Hunters Point, 72,650 lb of microscale ZVI wasmade into a 300-g/L slurry in tap water and was similarly injected by pneumaticfracturing into a region of more dilute contamination next to the DNAPL source.The iron-to-soil ratio achieved was 0.001. After ZVI injection, ORP dropped to

ii


5/54

below 400 mV in one well, but was between 200 mV and 400 mV in otherwells. Compared to the first study, ORP started rebounding more quickly afterthe second injection. Persistence of DCE and VC in the treatment zone after thesecond injection study is another indicator that insufficient iron may have beeninjected to generate the strongly reducing conditions necessary to stimulate themore efficient abiotic (beta-elimination) reactions that were created in the first

study. However, the mildly reducing conditions generated were sufficient topromote hydrogenolysis and anaerobic biodegradation of TCE.

At NAS Jacksonville, 300 lb of BNP from PARS Environmental Inc. was madeinto a 4.5- to 10-g/L iron slurry with water from an extraction well and injectedinto the subsurface by a combination of direct push and closed-loop recirculationwells. After injection, groundwater ORP dropped to below 200 mV, but pHremained relatively constant. The levels ofcis-DCE rose significantly, indicatingthat anaerobic biodegradation and hydrogenolysis were significantly stimulated,but aquifer conditions may not have been reducing enough to stimulate abioticreduction (beta-elimination) of TCE and other CVOCs. Either the iron waspartially passivated before injection when mixing with relatively high volume of

oxygenated water to form the injection slurry or the iron-to-soil ratio was nothigh enough to generate the strongly reducing conditions necessary for abioticreduction (beta-elimination).

At NAES Lakehurst, 300 lb of BNP from PARS Environmental Inc. was madeinto a relatively dilute 2-g/L slurry with water from an extraction well (NorthernPlume) and from a fire hydrant (Southern Plume) and injected into the subsurfaceby direct push. After injection, there was no change in ORP and pH. In fact,ORP increased in some wells. Either the iron was passivated before injectionwhen mixing with a relatively high volume of oxygenated water to form theinjection slurry or the iron-to-soil ratio was not high enough to generate thereducing conditions necessary to stimulate either anaerobic microbial degradation

or abiotic reduction. TCE, cis-DCE, and VC levels gradually decreased in themonitoring wells over several weeks of monitoring. The increase in ORP and thedecrease in CVOC concentrations may be indicative of dilution of contaminationdue to the injection of 18,000 gallons of oxygenated water into the treatmentzone.

The primary lessons learned from the NZVI application at these three Navy sites are asfollows:

Creating a reducing environment (ORP < 400 mV) strong enough to generatefaster abiotic reactions should be the main objective of NZVI treatment.

Abiotic (beta-elimination) reactions result in faster dehalogenation of CVOCsand minimal production of partially dechlorinated byproducts, such as cis-1,2-DCE.

When insufficient ZVI mass is injected, mildly reducing conditions (ORP


6/54

stimulating biodegradation alone may not be enough to justify the use of arelatively higher-priced reagent, such as NZVI.

Enough ZVI mass should be injected to lower the ORP below 400 mV. Fromthese demonstrations, an iron-to-soil ratio of 0.004 appears to be essential togenerate the required ORP. Injected ZVI mass should not be based on stoichoi-

metric relationships with the contaminant mass. Contaminant mass estimates,especially in DNAPL source zones, tend to be notoriously inaccurate and, in anycase, do not appear to directly drive iron requirements. Excess iron may have tobe injected to achieve an iron-to-soil ratio of 0.004 in the target zone afteraccounting for some migration of iron outside the target region.

In general, NZVI delivery mechanisms that minimize the volume of waterinjected along with the iron are preferable to methods that depend on largervolumes of water. Water from most sources contains oxygen and other oxidizedspecies that may passivate the iron during injection. If larger volumes of waterhave to be used, the water should be de-oxygenated first. However, otheroxidized species, such as nitrates and sulfates, may still persist and react with the

iron.

Larger ZVI particles are likely to be less reactive compared with true NZVIparticles. However, micron-sized or granular ZVI particles also may be lessprone to passivation during handling, compared with true NZVI particles. Thereactivity, stability, and cost of different-sized ZVI particles are all factors in theselection of a suitable treatment strategy.

Short-term monitoring of the treatment zone should demonstrate a congruence intrends among parent compounds (e.g., TCE), byproducts (e.g., cis-1,2-DCE), andother indicator parameters (e.g., ORP, pH, etc.). All these parameters shouldpreferably indicate strongly reducing conditions and abiotic (beta-elimination)

reactions.

Long-term monitoring of the treatment zone is essential until it is demonstratedthat the decline in parent compounds (e.g., TCE) and byproducts (e.g., cis-1,2-DCE) persists after ORP has rebounded to pre-treatment levels (that is, after theZVI is depleted). Only then can it be determined how much, if any, DNAPLmass truly remains in the treatment zone.

The results of this review indicate that NZVI injection is a promising option for treatment ofsource zones. Some additional long-term monitoring at all three sites may be worthwhile to evaluate thereasons for the differences in performance and to verify how much residual DNAPL there may be left atthese sites. Also of interest would be to see whether the downgradient plume has been weakened enough

by the source removal action for natural attenuation to be a viable long-term option or whether additionalNZVI injection is required.

iv


7/54

CONTENTS

EXECUTIVE SUMMARY .......................................................................................................................... iiFIGURES.....................................................................................................................................................vi TABLES .....................................................................................................................................................viiACRONYMS AND ABBREVIATIONS..................................................................................................viii

Section 1.0: INTRODUCTION ................................................................................................................... 11.1 Report Organization.................................................................................................................11.2 ZVI and NZVI Technologies Descriptions..............................................................................2

1.2.1 FEROXSM ....................................................................................................................3 1.2.2 Bimetallic Nanoscale Particles....................................................................................31.2.3 PARS Nanoscale Zero-Valent Iron (NZVI)................................................................31.2.4 Price of NZVI and its Variations.................................................................................4

Section 2.0: TECHNOLOGY IMPLEMENTATION.................................................................................. 52.1 Hunters Point ........................................................................................................................... 5

2.1.1 Site Description........................................................................................................... 5

2.1.2 Technology Implementation ....................................................................................... 72.1.3 Performance Evaluation Approach ............................................................................. 72.1.4 Technology Performance (Results).............................................................................82.1.5 Cost Summary...........................................................................................................132.1.6 Regulatory Issues ...................................................................................................... 132.1.7 Discussion ................................................................................................................. 13

2.2 Second Study at Hunters Point Shipyard...............................................................................142.2.1 Technology Implementation ..................................................................................... 142.2.2 Performance Evaluation Approach ........................................................................... 162.2.3 Technology Performance (Results)...........................................................................162.2.4 Discussion ................................................................................................................. 19

2.3 Naval Air Station Jacksonville .............................................................................................. 19

2.3.1 Site Description.........................................................................................................192.3.2 Technology Implementation ..................................................................................... 212.3.3 Performance Evaluation Approach ........................................................................... 232.3.4 Technology Performance (Results)...........................................................................232.3.5 Cost Summary...........................................................................................................272.3.6 Regulatory Issues ...................................................................................................... 282.3.7 Discussion ................................................................................................................. 28

2.4 Naval Air Engineering Station Lakehurst..............................................................................282.4.1 Site Description.........................................................................................................282.4.2 Technology Implementation ..................................................................................... 322.4.3 Performance Evaluation Approach ........................................................................... 322.4.4 Technology Performance .......................................................................................... 32

2.4.5 Cost Summary...........................................................................................................342.4.6 Regulatory Issues ...................................................................................................... 342.4.7 Discussion ................................................................................................................. 34

Section 3.0: DISCUSSION........................................................................................................................383.1 Technical Performance .......................................................................................................... 383.2 Cost Comparison ................................................................................................................... 403.3 Regulatory Issues...................................................................................................................423.4 Summary................................................................................................................................42

v


8/54

Section 4.0: REFERENCES ...................................................................................................................... 44

FIGURES

Figure 1-1. Abiotic Reduction of TCE by ZVI......................................................................................... 2Figure 2-1. Demonstration Site RU-C4, Hunters Point ............................................................................ 5Figure 2-2. Pre-Injection Baseline TCE Contaminant Plume................................................................... 6Figure 2-3. Pneumatic Fracturing System ................................................................................................ 8Figure 2-4. ORP Observed in Monitoring Wells IR28MW362F, IR28MW211F, and

IR28MW939F........................................................................................................................ 9Figure 2-5. pH Observed in Monitoring Wells IR28MW362F, IR28MW211F, and IR28MW939F....... 9Figure 2-6. TCE, cis-1,2-DCE, and Vinyl Chloride Concentrations in Source Area

Monitoring Well IR28MW362F .......................................................................................... 11Figure 2-7. TCE, cis-1,2-DCE, and Vinyl Chloride Concentrations in Source Area

Monitoring Well IR28MW939F .......................................................................................... 11Figure 2-8. TCE, cis-1,2-DCE, and Vinyl Chloride Concentrations in Source Area

Monitoring Well IR28MW211F .......................................................................................... 12Figure 2-9. TCE Concentrations in Source Area Monitoring Well IR28MW211F

from Long-Term Monitoring ............................................................................................... 12Figure 2-10. Pre-Injection Baseline TCE Contaminant Plume................................................................. 15Figure 2-11. ORP Observed in Monitoring Wells IR28MW310F, IR28MW409, and

IR28MW410A .....................................................................................................................16Figure 2-12. pH Observed in Monitoring Wells IR28MW310F, IR28MW409, and IR28MW410A ...... 17Figure 2-13. TCE, cis-1,2-DCE, and Vinyl Chloride Concentrations in Source Area Monitoring

Well IR28MW310F .............................................................................................................17Figure 2-14. TCE, cis-1,2-DCE, and Vinyl Chloride Concentrations in Source Area Monitoring

Well IR28MW409................................................................................................................ 18Figure 2-15. TCE, cis-1,2-DCE, and Vinyl Chloride Concentrations in Source Area Monitoring

Well IR28MW410A............................................................................................................. 18Figure 2-16. Site H1K, NAS Jacksonville................................................................................................ 20Figure 2-17. Baseline Contaminant Levels and TCE Contours................................................................ 22Figure 2-18. TCE, cis-1,2-DCE, and Vinyl Chloride Concentrations in Source Area

Monitoring Well H10MW37................................................................................................ 23Figure 2-19. TCE, cis-1,2-DCE, and Vinyl Chloride Concentrations in Source Area

Monitoring Well H10MW34................................................................................................ 24Figure 2-20. TCE, cis-1,2-DCE, and Vinyl Chloride Concentrations in Source Area

Monitoring Well H10MW32................................................................................................ 24Figure 2-21. ORP Observed in Monitoring Wells H10MW32, H10MW34, and H10MW37.................. 25Figure 2-22. pH Observed in Monitoring Wells H10MW32, H10MW34, and H10MW37..................... 26Figure 2-23. NAES Lakehurst Site Location Map.................................................................................... 29

Figure 2-24. Total VOC Concentrations in Monitoring Wells Baseline (Northern Plume) ..................... 30Figure 2-25. Total VOC Concentrations in Monitoring Wells Baseline (Southern Plume) ..................... 31Figure 2-26. TCE, PCE, DCE, and VC Concentrations in Source Area Monitoring Well MW-1........... 33Figure 2-27. TCE, PCE, DCE, and VC Concentrations in Source Area Monitoring Well MW-2........... 33Figure 2-28. ORP Levels at NAES Lakehurst .......................................................................................... 35Figure 2-29. pH Levels at NAES Lakehurst............................................................................................. 36

vi


9/54

TABLES

Table 2-1. Cost Summary for Hunters Point ZVI Demonstration ............................................................. 13Table 2-2. Key Groundwater Parameters in Treatment Zone Wells.......................................................... 26Table 2-3. Key Groundwater Parameters in Extraction Wells................................................................... 27Table 2-4. Costs for NZVI Demonstration at NAS Jacksonville............................................................... 27

Table 2-5. Costs for NZVI Demonstration at NAES Lakehurst ................................................................ 37Table 3-1. Key Results of the NZVI Demonstrations................................................................................ 38Table 3-2. Cost Breakdown for the Hunters Point, NAS Jacksonville, and NAES Lakehurst Sites ......... 41

vii


10/54

ACRONYMS AND ABBREVIATIONS

bgs below ground surfaceBNP bimetallic nanoscale particles

CERCLA Comprehensive Environmental Response, Compensation, and Liability Act

CVOC chlorinated volatile organic compound

DCA dichloroethaneDCE dichloroetheneDNAPL dense, nonaqueous-phase liquidDO dissolved oxygenDPT direct-push technology

FDEP Florida Department of Environmental Protection

HPS Hunters Point Shipyard

MW monitoring well

NA not available NAES Naval Air Engineering Station NAS Naval Air Station ND not detected NZVI nanoscale zero-valent iron

ORP oxidation-reduction potential

PCE tetrachloroethenePRB permeable reactive barrier

psi(g) pounds per square inch (gage)

RCRA Resource Conservation and Recovery Act

TCA trichloroethaneTCE trichloroetheneTOC total organic carbon

U.S. EPA (United States) Environmental Protection AgencyUST underground storage tank

VC vinyl chloride

VOC volatile organic compound

ZVI zero-valent iron

viii


11/54

Section 1.0: INTRODUCTION

This cost and performance report is a compilation of technical and performance data fromthree recent Navy demonstration projects involving the use of microscale or nanoscale zero-valent iron(NZVI) for treatment of dense, nonaqueous-phase liquid (DNAPL) source zones.

Zero-valent iron (ZVI) or elemental iron (Fe) is a strong reducing agent. In the past10 years, granular (coarse sand-sized) ZVI has been successfully used in permeable reactive barrier(PRB) applications to treat chlorinated volatile organic compounds (CVOC) in groundwater. Injectingfluidized NZVI particles into a contaminated source zone is an extension of this concept. Nanoscale ironparticles (typically between 50 to 300 nanometers in diameter) have surface areas that are up to severaltimes greater than larger-sized powders or granular iron. This characteristic makes NZVI particles muchmore reactive in a reduction-oxidation (redox) process. Microscale iron is a variant that consists ofmicron-scale particles that are coarser than NZVI particles but finer than granular iron. Concomitantly,the reactivity of microscale iron particles is expected to be less than that of NZVI, but greater than that ofgranular iron.

The Navy has conducted three field demonstration projects using various NZVI technologies

to determine their effectiveness in treating source zones contaminated primarily with CVOCs. Thedemonstration projects were conducted at three different Navy sites using various NZVI or microscaleiron technologies:

FEROXSM at the former Hunters Point Shipyard (Hunters Point), San Francisco,California.

PARS Environmental Inc.s NZVI Bimetallic Nanoscale Particle (BNP) at NavalAir Station (NAS) Jacksonville, Florida.

PARS Environmental Inc.s BNP at Naval Air Engineering Station (NAES)Lakehurst, New Jersey.

Each of these three projects is described in a separate application report. The current docu-ment is intended to be a comparative summary of the three projects and the demonstrated technologies.

1.1 Report Organization

This report is organized into the following sections:

Section 1.0: Introduction. This section provides the report framework, an introduction tothe NZVI technology, and the variations on this technology currently being demonstrated.

Section 2.0: Technology Implementation. This section provides details of the three

projects in details including technology performance and cost data. Regulatory issues and lessons learnedalso will be discussed briefly.

Section 3.0: Summary of Conclusions. This section compares and contrasts the threeprojects based on three categories: Technical Performance, Cost and Regulatory Issues, and LessonsLearned.

Section 4.0: References. This section lists the references used to prepare this report.

1


12/54

1.2 ZVI and NZVI Technologies Descriptions

ZVI is currently used primarily in PRBs for in situ remediation of chlorinated solvent plumes(ITRC, 2005; Gavaskar et al., 2002). Certain metals, such as chromium, in groundwater also have beentreated with ZVI in PRBs. These passive treatment walls for dissolved-phase contaminants are normallylocated in a plume to intercept groundwater flow prior to its migration off-site or towards potential

receptors. As the dissolved-phase contamination flows through the PRB, CVOCs are destroyed primarilyby abiotic reduction. Roberts et al. (1996) have proposed that a CVOC, such as trichloroethene (TCE), isdechlorinated by the two reactions shown in Figure 1-1. Most of the TCE is converted to ethene andchloride by beta-elimination reaction, which proceeds with the formation of short-lived intermediates,such as acetylene. A small portion of TCE decomposes by hydrogenolysis, a sequential reduction path-way that results in the formation of longer-lived intermediates, such as cis-1,2-dichloroethene (cis-1,2-DCE) and vinyl chloride (VC).

Hydrogenolysis

Beta-Elimination

Figure 1-1. Abiotic Reduction of TCE by ZVI(Permission has been received from the author to use the information from Roberts et al. 1996

Reductive Elimination of Chlorinated Ethylene by Zero-Valent Metals to create this figure.)

When enough ZVI is present, hydroxyl radicals are produced as water decomposes and thistends to increase pH. Native groundwater species, such as nitrate and sulfate, also are reduced. Calciumand magnesium tend to precipitate out as carbonates. Metals, such as chromium, are reduced to a loweroxidation state, at which they are less soluble and can be removed by precipitation.

The granular (coarse) nature of the ZVI particles used in PRBs ensures that adequatehydraulic flow capture is achieved and further downgradient migration of contamination emanating fromthe source is arrested. These passive treatment systems are simple and cost-effective to implement atrelatively shallow depths and for relatively low concentrations of CVOCs. However, challenges withimplementation and cost increase in deeper aquifers and at higher CVOC concentrations.

Because of its extremely small size and high surface area, NZVI is thought to be a moreeffective technology for remediation of source zones. An enlarged surface area allows the NZVI particlesto react at a much higher rate with CVOCs. This potentially improves remediation performance in high-concentration portions of the plume and in the source zone. Small particle size also allows much moremobility into the soil pores and NZVI can more easily be injected into shallow and deep aquifers than

2


13/54

granular iron. Ability to inject iron can be advantageous when contamination underlies a building. Inaddition, strongly negative redox conditions within the zone of treatment create conditions favorable foranaerobic microbial growth and enhance bioremediation of CVOCs (ITRC, 2005; Gu et al., 2002).

1.2.1 FEROXSM. The FEROXSM technology, patented by ARS Technologies Inc., involvesinjection of fluidized ZVI powder into the target zone of the subsurface. This method has the potential to

provide in situ treatment to both groundwater and soils contaminated with CVOCs and/or leachable heavymetals. The ZVI powder can be injected into the subsurface as a slurry or as a dry material. Nitrogen gasor compressed air is used as a carrier fluid. Pneumatic fracturing, a technique that injects gas into thesubsurface at low pressure and high volume to develop a network of fractures in the treatment zone, canbe used to promote movement of the ZVI slurry into the entire treatment zone. The iron powder used inthis technology is not strictly NZVI, as the particles are in the micron-size range.

According to ARS, the atomized multi-phase injection approach provides several key benefitsover conventional injection techniques including:

1. Aggressive mixing / recirculation maintains the iron powder in uniform suspension andallows for the reaction of the iron powder with water to be accelerated

2. Injection of NZVI slurry provides added moisture necessary for the reaction whenapplying in an unsaturated zone.

3. The injection approach allows the iron powder to be injected into the formation tosignificant radial distances using relatively low pressures (


14/54

environmentally friendly polymer coating to prevent particle accumulation and their adhesion to soilsurfaces. Consequently the product may be introduced into the subsurface by gravity feed injection. Thepolymer is non-toxic and is approved by the U.S. Food and Drug Administration as a food additive. TheNZVI particles dissolve in groundwater as ferrous iron (Fe+2). Depending on the pH and redox conditionsin groundwater that may change along the groundwater flow path, the resulting ferrous iron will formaggregates or flocs of iron oxyhydroxide. These solids are expected to be immobilized as surface coat-

ings on the soil surfaces in the aquifer.

1.2.4 Price of NZVI and its Variations. The price of NZVI has decreased in the past year due toa decrease in the price of raw materials used in nanoscale iron manufacturing, increased manufacturingcapacity, and an increased number of suppliers and vendors. Prices for 1,000-pound quantities or more ofNZVI vary from $31 per pound (unsupported, non-catalyzed nanoscale iron), $45 per pound (supported,non-catalyzed), and $66 per pound (supported, catalyzed), as quoted by PARS Environmental Inc. Theprice of the iron varies significantly depending on a number of factors including raw material cost,manufacturing cost, licensing fees, and other economic factors (such as supply and demand). As a resultof significant variability in the type of nanoscale iron and catalyst/support selected, NZVI products canvary significantly in physical-chemical characteristics and performance.

Current price quotes obtained from numerous NZVI vendors varied from $20 to $77 perpound depending on the quantity. Zloy, a product from OnMaterials, Inc., was recently quoted at $20 perpound plus additional shipping of $3 per pound. PolyMetallixTM, a product manufactured by CraneCompany and distributed by Nanitech, LLC, was recently quoted at a total delivered cost of $77 perpound for 300 lb; an increased quantity of 400 lb would have a slightly lower unit price of $72 per pound.RNIP, a product from Toda America, ranges from approximately $26 to $34 per pound (greater than10,000-lb quantity to less than 100 lb, respectively), plus additional shipping and handling charges.Presently, ARS Technologies, Inc. offers uncatalyzed microscale ZVI for a price between $1.70 to1.00 per pound for quantities ranging from 1,000 to 1 million pounds, not including freight and tax(www.zerovalentiron.com).

4
http://www.zerovalentiron.com/http://www.zerovalentiron.com/


15/54

Section 2.0: TECHNOLOGY IMPLEMENTATION

This section describes the NZVI treatment conducted in CVOC source zones at three Navysites. The different site conditions and application challenges illustrate the types of considerations drivingthe use of this technology.

2.1 Hunters Point

2.1.1 Site Description. Hunters Point is situated on a long promontory, located in the southeasternportion of San Francisco County and extends eastward into San Francisco Bay as shown in Figure 2-1.From 1869 through 1986, Hunters Point operated as a ship repair, maintenance, and commercial facility.In 1991, the Navy designated Hunters Point for closure under the federal Base Closure and RealignmentAct. Hunters Point was divided into six separate geographic parcels (Parcels A through F) to facilitate theclosure process. This ZVI demonstration was performed at Site RU-C4 in Parcel C, which is located inthe eastern portion of Hunters Point.

Figure 2-1. Demonstration Site RU-C4, Hunters Point

Two aquifers and one water-bearing zone have been identified at Hunters Point: theA-aquifer, the B-aquifer, and the bedrock water-bearing zone. Groundwater flow patterns are complexdue to heterogeneous hydraulic properties of the fill materials and weathered bedrock, tidal influences,effects of storm drains and sanitary sewers, and variations in topography and drainage. RU-C4 hydro-geology is characterized by shallow bedrock with a rolling and uneven surface overlain predominantly byartificial fill material of variable hydraulic conductivity. The A-aquifer is unconfined and directly over-lies the bedrock water-bearing zone. In the western portion of Parcel C where bedrock is present atshallow depths, B-aquifer zones are isolated and mostly absent. During this demonstration, groundwater

5


16/54

levels were consistent with previous measurements at this site and ranged from an average of 6.8 ft belowground surface (bgs) prior to injection to an average of 6.2 ft bgs post-injection. At RU-C4, based on slugtests, hydraulic conductivity in the A-aquifer ranged from 26.6 to 43 ft per day, and hydraulic conductiv-ity in the bedrock water-bearing zone ranged from 5.2 102 to 40 ft per day. Groundwater gradients aregenerally flat, with a historically measured gradient of about 0.0025 to the south-southwest (Tetra TechInc., 2003).

The contaminant plume at RU-C4 consists of chlorinated solvents, primarily TCE, in shallowgroundwater beneath the northern portion of Building 272. Possible sources of chlorinated solvents atRU-C4 include (1) a former underground storage tank (UST) used for waste oil storage, immediatelynorth of Building 272, and the associated floor drain and underground piping inside of Building 272;(2) a grease trap, immediately north of Building 272 (east of the former UST), and the associated cleanoutand underground piping inside Building 272; and (3) five steel dip tanks at a former paint shop in thesouthwestern portion of Building 281.

Groundwater contaminant characterization conducted before this demonstration indicated thatTCE was present at high concentrations in shallow groundwater in an isolated area beneath the north-eastern portion of Building 272 (PRC et al., 1997; Tetra Tech Inc., 2003). These concentrations sug-

gested the likely presence of DNAPL in a small portion of the TCE plume. However, DNAPL was notobserved during the baseline sampling or in subsequent groundwater monitoring events. Results frombaseline sampling were generally consistent with previously reported concentrations and further refinedthe spatial extent of the plume, as shown on Figure 2-2, which presents horizontal baseline TCEisoconcentration contours.

Figure 2-2. Pre-Injection Baseline TCE Contaminant Plume

6


17/54

2.1.2 Technology Implementation. Feroxsm microscale ZVI was used for the demonstration atHunters Point. Feroxsm injections were conducted in four open boreholes. Figure 2-2 shows the locationsof the four injection boreholes and surrounding monitoring wells. The four injection boreholes were eachdrilled to a depth of 32 ft bgs. Temporary 4-inch-diameter steel casings with disposable tips were thenpushed to depth using a direct-push rig to prevent caving prior to injection. Injection boreholes weredrilled to a depth below where DNAPL would potentially be observed, and injections were performed

from the bottom up to minimize the potential risk of displacing DNAPL horizontally or downward intothe bedrock water-bearing zone.

The design dosage of ZVI powder was 16,000 lb. This dosage was based on (1) the estimatedmass of TCE, which makes up most of the total CVOCs; (2) the estimated mass of soil within the treat-ment zone; and (3) the mass ratios of iron-to-TCE and iron-to-soil. The design dosage was calculatedusing these two mass ratios, as well as safety factors, to account for fluctuations in historic TCE concen-trations, unknown sources, and less than ideal distribution of the ZVI powder. The mass of TCE withinthe treatment zone was estimated to be about 14 lb. Successful injection and placement of 16,000 lb ofZVI would achieve an iron-to-TCE mass ratio of about 1,100. In general, an iron-to-soil mass ratio of0.004 is necessary to achieve a sufficient reductive environment for the abiotic degradation of TCE.Based on an estimated dimension for the treatment zone of about 900 ft2 by 22 ft in thickness, the mass of

soil within the treatment zone was estimated to be about 1,980,000 lb. As a result, successful emplace-ment of 16,000 lb of ZVI was designed to achieve an iron-to-soil mass ratio of about 0.008. After theZVI injection were completed, the vendor estimated that a radius of influence of 15 ft was achieved andbased on this distribution the ZVI-affected region was estimated to encompass 4,544,100 lb of soil, whichis approximately twice the designed treatment mass of soil. Therefore, the actual iron-to-soil ratioprobably was closer to 0.004 than to 0.008.

The injection process integrated pneumatic fracturing and Feroxsm delivery into one process.Nitrogen gas was used as both a fracturing and injection fluid. Injections were conducted sequentially ineach of the four boreholes at 3-ft intervals, starting at a depth of 30 ft bgs and proceeding upward to atleast 10 ft bgs. This series of injections was expected to vertically cover the zone from 32 ft bgs to about7 ft bgs (the approximate water table), or the zone where significant concentrations of CVOCs had been

measured. Figure 2-3 is a schematic diagram of the pneumatic fracturing and Feroxsm

injection process.The pneumatic fracturing system included a specialized injection module that reduced and regulated theflow of compressed nitrogen to pressures used for both fracturing and injection. Fracturing pressuresranged from 55 to 230 pounds per square inch gauge (psig) (ARS Technologies Inc., 2003). ZVI powderand potable water were combined at a ratio of 1 kilogram of ZVI powder to 1 gallon of water, to createthe ZVI slurry. Subsurface pressures during the injection process ranged from 40 and 180 psig (ARSTechnologies Inc., 2003).

Initial responses to injections indicated that gas dissipated slowly through the formation.Most injections were conducted using pulses of nitrogen, instead of steady flows, to minimize the amountof nitrogen introduced to the injection and to prevent excessive buildup of pressure and surface heave.The quantity of ZVI injected within each interval varied based on the duration and number of injections in

each borehole.

2.1.3 Performance Evaluation Approach. The Navy conducted four rounds of groundwatersampling to evaluate the effectiveness of the ZVI injections. A baseline round was conducted prior to theinjections, and three post-injection demonstration rounds were conducted at 2, 6, and 12 weeks after theinjections. Eighteen monitoring locations were selected for sampling to represent the areas within,upgradient, cross-gradient, and downgradient of the expected treatment zone. These wells are screened inthe zone of vertical coverage of the ZVI injections (7 to 32 ft bgs). One well located within the horizontalextent of the treatment zone, but below the vertical coverage of the ZVI injections, also was selected for

7


18/54

Figure 2-3. Pneumatic Fracturing System

sampling to represent the area below the treatment zone. Tetra Tech Inc. measured groundwater samplesfor time-sensitive parameters in the field. Additional samples were sent to a laboratory and analyzed forvolatile organic compounds (VOCs), dissolved hydrogen gases, dissolved arsenic, dissolved iron, totaliron, dissolved manganese, chloride, alkalinity, and nitrate.

The performance evaluation was based on the percent reduction in contaminant level from thebaseline assessment to post-injection. Percent reductions were calculated based on mean concentrationobserved over the three post-injection sampling rounds. Additional long-term monitoring was conductedin select wells in the treatment area approximately one year after the ZVI injections.

2.1.4 Technology Performance (Results). Measurement of various groundwater parameters thatindicate the presence of iron or the occurrence of dechlorination reactions were measured near injectionpoints. Oxidation-reduction potential (ORP) results less than 200 mV were observed at distances of15 ft or less from each injection borehole. In many wells, ORP declined to below 400 mV, which isindicative of an environment strongly reducing enough for abiotic reactions to occur.

Because aquifer parameters such as ORP could change over time, temporal post-injectionaverages may not be a suitable way of representing the effects of the ZVI, as has been done in the imple-mentation report. For example, when the ORP observed during each individual post-injection monitoringevent in monitoring wells IR28MW362F, IR28MW211F, and IR28MW939F is plotted in Figure 2-4, atleast two wells in the treatment zone shows signs of a gradual rebound, indicating that the iron in thevicinity of this well is dissipating or is being passivated. Temporal means tend to suppress these gradualchanges in the redox environment of the aquifer.

8


19/54

-600

-500

-400

-300

-200

-100

0

100

200

300

S

ep-02

D

ec-02

Mar-03

Jun-03

Oct-03

Jan-04

Apr-04

A

ug-04

N

ov-04

Feb-05

Date

ORP(mV)

IR28MW362F

IR28MW211F

IR28MW939F

Figure 2-4. ORP Observed in Monitoring Wells IR28MW362F, IR28MW211F, and IR28MW939F

An increase in pH would also be indicative of treatment occurring. The pH increased afterinjection at all but two locations: monitoring well IR28MW934F5, located 13.1 ft from the nearestinjection point and IR28MW351F, located 4.8 ft from the nearest injection borehole. However,IR28MW934F5 was screened below the targeted treatment zone at 51 to 59 ft bgs. All other wells insidethe treatment zone within 15 ft of an injection point showed a pH increase of about 1 to 2 pH units. ThepH values observed in monitoring wells IR28MW362F, IR28MW211F, and IR28MW939F are shown onFigure 2-5. The ORP and pH measurements in these wells indicate that strongly reducing conditions

necessary for abiotic reduction of TCE were achieved in most locations.

0

2

4

6

8

10

12

IR28MW362F

IR28MW211F

IR28MW939F

pH

Dec-02

Jan-03

Mar-03

Mar-03

Aug-04

Oct-04

Nov-04

Jan-05

Figure 2-5. pH Observed in Monitoring Wells IR28MW362F, IR28MW211F, and IR28MW939F

9


20/54

Changes in concentrations of dissolved gases (ethane, ethene, and hydrogen) provided asimilar indication of the extent of the treatment zone. Increases in ethane and ethene concentrations wereobserved at most locations 15 ft or less from the nearest injection borehole. Dissolved hydrogen resultsdid not indicate the extent of the treatment zone because it was not detected during the baseline samplinground and was detected at only three locations after the injection process was completed. However,hydrogen is quickly consumed by microbes and even sporadic detections of hydrogen indicate that

reducing conditions have been generated.

Although increases in chloride concentrations could be expected to result from the ZVIinjection, chloride concentrations actually decreased at all but four locations varying in distance frominjection. Because background chloride concentrations in groundwater are relatively high, it is likely thatthe variations in background concentrations outweighed any chloride production that resulted from treat-ment. Alkalinity levels decreased at all locations within the treatment zone and did not change signifi-cantly outside the treatment zone. The data discussed above support the conclusion that the treatmentzone extended to distances of at least 15 ft from the point of injection, covering an area of about 1,818 ft2.

Groundwater samples were collected for analysis of CVOCs before and after the injectionprocess. Results of these samples were used to estimate percent reduction of the following CVOCs of

concern: four chlorinated ethenes (tetrachloroethene [PCE], TCE, 1,2-DCE, and VC), total chlorinatedethenes, and two additional CVOCs (chloroform and carbon tetrachloride). Percent reduction was calcu-lated for these compounds by comparing concentrations within the treatment zone before and after ZVIinjection. Post-injection concentrations were represented by the average concentration measured duringthe three post-injection sampling rounds. The percent reduction calculated based on the arithmetic meanof concentrations within the treatment zone is considered more meaningful because the calculated valuesaccount for both decreases and increases in concentrations at individual monitoring locations. However,care has to be taken to average concentrations measured over multiple monitoring events, as this tends tosuppress temporal trends that may be important for evaluating remediation effectiveness.

Analytical results for the 10 monitoring wells located within 15 ft of injection points wereused to estimate percent reduction of CVOCs within the treatment zone. The highest pre-injection

concentration (88,000 g/L) of TCE was observed at injection borehole F2, which was later converted tomonitoring well IR28MW362F. Post-injection results at this borehole averaged at a concentration of31 g/L, reflecting a percent reduction for this individual location of 99.96%, as shown on Figure 2-6. Inthe treatment zone, the overall mean pre-injection concentration of TCE was 27,000 g/L and the overallmean post-injection concentration was 220 g/L. These concentrations represent an overall percentreduction of TCE within the treatment zone of 99.2%. Reduction of TCE to ethene and chloride was con-siderable and no significant formation of intermediate degradation products (cis-1,2-DCE, trans-1,2-DCE,1,1-DCE, and VC) was observed. The overall reduction in dissolved concentrations as measured in themonitoring wells in the treatment zone for the CVOCs of concern were as follows: TCE (99.2%), PCE(99.4%), cis-1,2-DCE (94.2%), VC (99.3%), total chlorinated ethenes (99.1%), chloroform (92.6%), andcarbon tetrachloride (96.4%). CVOC concentrations in the weeks following injection for monitoringwells IR28MW939F and IR28MW211F are shown in Figures 2-7 and 2-8, respectively.

When the CVOC concentrations measured during each post-injection monitoring event areplotted (Figures 2-6 to 2-8), some interesting temporal trends are observed that are not evident in the post-injection mean concentrations. Anaerobic biostimulation is a desirable side effect of ZVI injection, butthe persistence ofcis-1,2-DCE in some wells indicates that some TCE DNAPL may still be present in thetreatment zone. In fact, in well IR28MW939F (Figure 2-7), there were signs that both cis-1,2-DCE andTCE levels are increasing gradually. It is interesting to note in Figure 2-4 that ORP levels in this well areshowing signs of a gradual rebound (increase) over time, as the iron loses some of its reactivity.Additional monitoring events are needed to evaluate the sustainability of these trends.

10


21/54

0

12,500

25,000

37,500

50,000

62,500

75,000

87,500

100,000

Nov-02

Dec-02

Dec-02

Jan-03

Feb-03

Feb-03

Mar-03

Apr-03

Date

TCEConcentratio

n(g/L)

0

25

50

75

100

125

150

175

200

cis-1,2-DCEandVCConcentration(g/L)

TCE

cis-1,2-DCE

Vinyl Chloride

Figure 2-6. TCE,cis-1,2-DCE, and Vinyl Chloride Concentrations in Source AreaMonitoring Well IR28MW362F

0

2,000

4,000

6,000

8,00010,000

12,000

14,000

16,000

18,000

20,000

Nov-02

Dec-02

Dec-02

Jan-03

Feb-03

Feb-03

Mar-03

Apr-03

Date

TCEConcen

tration(g/L)

0

50

100

150

200250

300

350

400

450

500

cis-1,2-DCEandVCConcentrations(g/L)

TCE

cis-1,2-DCEVinyl Choride


In monitoring well IR28MW211F (Figure 2-8), cis-1,2-DCE levels, after an initial sharpdecline two weeks after injection, show signs of increasing in subsequent weeks. This indicates that asthe direct influence (abiotic reduction) of the ZVI wanes, anaerobic microbial stimulation kicks in andcontinues to biodegrade any residual TCE. Subsequent longer-term monitoring in well IR28MW211F

11


22/54

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

80,000

Nov-02

Dec-02

Dec-02

Jan-03

Feb-03

Feb-03

Mar-03

Apr-03

Date

TCEConcentration(g/L)

0

25

50

75

100

125

150

175

200


TCEcis-1,2-DCE

Vinyl Chloride


0

200

400

600

800

1,000

1,200

1,400

1,600

1,800

Feb-03

Jun-03

Sep-03

Dec-03

Apr-04

Jul-04

Oct-04

Jan-05

May-05

Date

Con

centration(g/L)

TCE

cis-1,2-DCE

Vinyl Chloride

Figure 2-9. TCE Concentrations in Source Area Monitoring Well IR28MW211Ffrom Long-Term Monitoring

shows a continued decrease in TCE concentrations (Figure 2-9). In the longer term, cis-1,2-DCE levelscontinued to increase until a second ZVI injection was conducted in a neighboring area; beyond this time,cis-1,2-DCE levels declined, whereas VC levels increased slightly. This indicates that hydrogenolysisand biodegradation of CVOCs is continuing, and the byproducts of TCE degradation themselves aredegrading. It would be interesting to continue monitoring the treatment zone wells until ORP returns topre-treatment levels. Once ORP returns to pre-treatment levels, and if TCE does not rebound signifi-cantly, that would be a strong indicator that DNAPL destruction is mostly complete.

12


23/54

Because a reduction in VOC concentrations also could be interpreted as the result of plumedisplacement rather than treatment, sampling locations outside the treatment zone were monitored forpotential increases in contaminant concentrations. At locations outside of the treatment zone, TCE ingroundwater generally remained the same, either slightly increasing or decreasing. Monitoring wellIR28MW360F was an exception, having TCE concentrations decrease from a baseline of 7,400 g/L to apost-injection average of 640 g/L. Increases in ethane and ethene concentrations also were observed at

well IR28MW360F; however, the overall increase in ORP and the distance of this well from the nearestinjection borehole suggest that this location was outside the treatment zone. Excluding IR28MW360F,the mean concentration of TCE in groundwater at locations outside the treatment zone increased slightlyafter injection, by 15 g/L. Because the increase was minor in comparison with the decrease of the meanconcentration of TCE within the treatment zone, it can be concluded that the plume was not being signifi-cantly displaced.

2.1.5 Cost Summary. Table 2-1 summarizes costs of the field-scale technology demonstration.The table does not include costs for the demonstration plans including work plans, health and safety plan,and demonstration-derived waste plan, project management, and contractor and health and safetyoversight.

Table 2-1. Cost Summary for Hunters Point ZVI Demonstration

Category/Item Cost % of Total Cost

Mobilization $31,200 10.8

Equipment and Supplies(ZVI Cost = $32,500)

$99,900 34.5

Labor $38,900 13.5

Drilling Services $22,800 7.9

Sampling and Analysis includingDerived Waste Analysis and Disposal

$92,600 32

Other Miscellaneous costs $3,900 1.3

Total Cost of Demonstration $289,300 100

2.1.6 Regulatory Issues. No significant permitting or regulatory issues were identified as aconcern during the demonstration. Long-term groundwater monitoring plans were not included in thescope of this project.

2.1.7 Discussion. Some of the conclusions and lessons drawn from the demonstration included:

Dissolved TCE levels declined sharply in all monitoring wells in the treatmentzone, without any significant formation ofcis-1,2-DCE and VC. This indicatesthat the ZVI injection successfully created the conditions necessary for abioticreduction of CVOCs by beta-elimination reactions.

Sharp declines in ORP and noticeable increases in pH support the contention thatthe strongly reducing conditions suitable for abiotic reduction of CVOCs werecreated. Native chloride levels are relatively high at this site and chlorideformation was not a good indicator of dechlorination of CVOCs.

Injecting a ZVI mass in excess of contaminant stoichiometry was necessary tobring about significant abiotic reduction of CVOCs. ZVI mass should be suffi-cient to generate the sharply reducing conditions necessary to induce abiotic

13


24/54

degradation reactions (primarily beta-elimination) that do not result in generationof partially dechlorinated byproducts.

Long-term performance measures need to be included in the monitoring plan.Even with excess iron, the DNAPL source could be temporarily suppressed, butrebound of dissolved CVOCs could eventually occur. ORP levels should

continue to be monitored. If CVOC levels remain low after ORP rebound to pre-treatment levels has occurred, then source treatment can be said to be complete.

Pneumatic fracturing combined with liquid atomization injection of the ZVIslurry was successful in distributing ZVI through most of the target treatmentzone. Slow nitrogen distribution through the formation at Hunters Point led thevendor to use pulses of nitrogen rather than a steady flow to distribute the ZVI.Pulsing helped to prevent excessive pressure buildup and surface heave.

Injecting at shallow depths may lead to nitrogen and slurry seeping up to theground surface. Switching to direct hydraulic pumping may reduce the potentialfor seeping to the ground surface and the risk of contaminant vapors escaping

from the subsurface. Some minor heaving of the concrete floor was observedduring the demonstration, which generally occurred during shallow injections.However, residual heave up to 1 inch was observed after the injections. Extraprecautions may be required if this technology is applied under buildings.

2.2 Second Study at Hunters Point Shipyard

A second field treatability study was performed at the RU-C4 site using the same technologyto determine its effectiveness at treating a larger plume that contained lower contaminant concentrations(ITSI, 2005). Prior to implementing the treatability study, additional delineation of the groundwatercontaminant plume was necessary. Once the treatment zone was defined, injection well and monitoringwell locations were determined. A baseline groundwater sampling event was conducted. The contami-

nant plume which reflects the results from baseline sampling is shown in Figure 2-10.

2.2.1 Technology Implementation. Feroxsm injections were conducted in 13 open boreholes asshown on Figure 2-10. Each injection point contained between 4 and 12 intervals. 600 to 1,200 lb of ZVIwere injected into each depth interval. The design dosage of ZVI powder was 60,000 lb based on thesame assumptions used in the previous study at Hunters Point. As opposed to the first field study, inwhich an iron-to-soil ratio of 0.008 was targeted and a ratio of 0.004 achieved (due to a higher-than-expected radius of distribution of the ZVI), in the second study, an iron-to-soil ratio of 0.004 was targetedand a 0.001 ratio was achieved. Some daylighting of ZVI to the ground surface was observed and somewells outside the target treatment region showed signs of ZVI influence, indicating some migration ofZVI outside the target zone. The actual amount of ZVI injected into the subsurface was 72,650 lb. Theinjection process used was the same Feroxsm delivery methodology that was implemented in the first

study. Injections were conducted sequentially in each of the 13 boreholes at 3- to 4-ft intervals, starting atthe deepest interval and proceeding upward.

The total cost for the treatability study implementation was $1,390,000. The total costincluded $770,000 for materials, equipment, field labor for the ZVI injection, and waste characterizationand disposal; $452,000 for baseline and post-injection groundwater sampling and laboratory analysis; and$168,000 for project management, data management, and reporting. The total treatment volume for thetreatability study was estimated to be 27,778 yd3. Therefore, the cost to treat one cubic yard of soilwas $50.

14


25/54

Figure 2-10. Pre-Injection Baseline TCE Contaminant Plume

15


26/54

2.2.2 Performance Evaluation Approach. Four rounds of groundwater sampling were conductedto evaluate the effectiveness of the ZVI injections. A baseline round was conducted prior to the injec-tions, followed by three post-injection demonstration rounds. Nineteen monitoring locations wereselected for sampling to represent the areas within, near the perimeter, and outside of the expected treat-ment zone. These wells are screened in the zone of vertical coverage of the ZVI injections (up to 60 ftbgs). Samples were sent to an off-site laboratory and analyzed for VOCs, dissolved hydrocarbon gases,

total and dissolved iron, and alkalinity. The performance evaluation was based on the percent reductionin contaminant level from the baseline assessment to post-injection monitoring. Percent reductions werecalculated by site representatives based on mean concentrations observed over the three post-injectionsampling rounds. As mentioned in Section 2.1.4, temporal means may tend to suppress important timetrends in the monitored parameters and individual data points also need to be examined.

2.2.3 Technology Performance (Results). ORP declined sharply in several treatment zone wells,thus indicating relatively good ZVI distribution (Figure 2-11), and the pH of the groundwater in thetreatment zone increased significantly (to levels of 10 or above in some wells) (Figure 2-12). The declinein ORP in the second study was not as sharp as the decline in the first study (see Figures 2-11 and 2-4).In the first study, ORP declined to below 500 mV in most wells, whereas in the second study, ORPdeclined to below 400 mV in only one well and below 300 mV in another well. In most other

treatment zone wells, ORP declined to approximately

200 mV.

The rebound in ORP also appears to have happened sooner in the second study, as seen inFigure 2-11. In fact, the ORP in monitoring well IR28MW410A is gradually approaching pre-treatmentlevels. Interestingly, ORP levels in several wells outside the treatment zone declined sharply, thus indi-cating that the ZVI had migrated to regions outside the target zone. This is borne out by a concomitantincrease in total iron levels in surrounding wells.

As seen in Figures 2-13 to 2-15, the decline in ORP was sufficient to cause sharp reductionsin TCE levels in several treatment zone wells. However, the decline in cis-1,2-DCE was not as sharp

-500

-400

-300

-200

-100

0

100

200

Aug-04

Aug-04

Sep-04

Sep-04

Oct-04

Nov-04

Nov-04

Dec-04

Jan-05

Jan-05

Date

ORP(mV)

IR28MW310F

IR28MW409

IR28MW410A

Figure 2-11. ORP Observed in Monitoring Wells IR28MW310F, IR28MW409, and IR28MW410A

16


27/54

0

2

4

6

8

10

12

IR28MW310F

IR28MW409

IR28MW410A

pH

8/10/2004

10/26/2004

11/29/20041/3/2005

Figure 2-12. pH Observed in Monitoring Wells IR28MW310F, IR28MW409, and IR28MW410A

0

200

400

600

800

1,000

1,200

Aug-04

Aug-04

Sep-04

Sep-04

Oct-04

Nov-04

Nov-04

Dec-04

Jan-05

Jan-05

Date

TCE

Concentration(g/L)

0

25

50

75

100

125

150

175

200

cis-1,2-DCE

andVCConcentration(g/L)

TCE

cis-1,2-DCE

Vinyl Chloride

Figure 2-13. TCE,cis-1,2-DCE, and Vinyl Chloride Concentrations in Source Area MonitoringWell IR28MW310F

(compared to the first injection study) and concentrations rebounded relatively quickly in some wells.VC concentrations also increased slightly in some wells. Monitoring well IR28MW310F shows the bestTCE, cis-DCE, and VC treatment performance, indicating strong abiotic (beta-elimination) reactions.In the other treatment zone wells, a probable scenario is that after some initial faster abiotic (beta-elimination) reactions, these regions are seeing more gradual biodegradation and hydrogenolysis of TCEand DCE. An increase in ethene levels indicates that, given enough time, CVOC dechlorination is goingto completion.

17


28/54

0

100

200

300

400

500

600

700

Aug-04

Aug-04

Sep-04

Sep-04

Oct-04

Nov-04

Nov-04

Dec-04

Jan-05

Jan-05

Date

TCEConcentration(g/L)

0

20

40

60

80

100

120

140

160

180

200

cis-1,2-DCEandVCConce

ntration(g/L)

TCE

cis-1,2-DCE

Vinyl Chloride

Figure 2-14. TCE,cis-1,2-DCE, and Vinyl Chloride Concentrations in Source AreaMonitoring Well IR28MW409

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

Aug-04

Aug-04

Sep-04

Sep-04

Oct-04

Nov-04

Nov-04

Dec-04

Jan-05

Jan-05

Date

TCEConcentratio

n(g/L)

0

50

100

150

200

250

300

350


TCE

cis-1,2-DCE

Vinyl Chloride

Figure 2-15. TCE,cis-1,2-DCE, and Vinyl Chloride Concentrations in Source Area MonitoringWell IR28MW410A

18


29/54

2.2.4 Discussion. Some of the conclusions and lessons drawn from the demonstration included:

The second ZVI injection study was targeted towards a larger region of the aqui-fer that contained lower levels of CVOC contamination compared to the first one.

An iron-to-soil ratio of 0.004 was targeted in the treatment zone, but was notachieved and may be attributed to the migration of substantial ZVI mass outsidethe target treatment zone. The resulting decline in ORP was not as sharp as thatexperienced in the first study (where an iron-to-soil ratio of 0.008 was targetedand 0.004 was achieved). An iron-to-soil ratio of 0.001 appears to have beenachieved in the target zone in the second study.

TCE was reduced rapidly in the treatment zone wells, as was DCE. However,DCE is already showing signs of rebound in several wells and TCE itselfappeared to be rebounding in one well, MW410A (the well in which ORP is alsoshowing a rebound to pre-treatment levels). This indicates that dissolved-phaseTCE was treated in the short-term, but sorbed TCE may gradually show up asdissolved-phase in the monitoring wells.

These results indicated that TCE and DCE initially were reduced by strongabiotic reactions in some portions of the target treatment zone. However, insome portions of the treatment zone, TCE and DCE are degrading mostly byslower biodegradation or hydrogenolysis reactions.

2.3 Naval Air Station Jacksonville

2.3.1 Site Description. NAS Jacksonville is located in Duval County, Florida and has been usedfor Navy operations since 1940. The demonstration site, H1K (Tetra Tech, Inc., 2005), was located in theinterior portion of the NAS Jacksonville and contained two USTs, Tank A and Tank B. The USTspreviously received waste solvents and other substances from a wash rack, manhole, and other operations.

The tanks and associated pipelines were removed and capped in 1994 and are suspected to be the sourceof contamination. Cleanup of H1K is managed under the Comprehensive Environmental Response,Compensation, and Liability Act (CERCLA) program, and the groundwater monitoring program ismanaged under the Resource Conservation and Recovery Act (RCRA) program. Based on confirmationsoil samples collected in 1995, the tank and pipeline removal achieved clean closure for unsaturated soils.Elevated concentrations of TCE and 1,1,1-trichloroethane (TCA) in groundwater are centered in the TankA area. The site location is shown in Figure 2-16.

Geologic borings indicate that the unsaturated zone at the site appears to be fairly uniformfine to medium grained sand and sandy fill. A thin layer of clayey sand and/or silty sand is located at andjust below the water table between 6 and 12 ft bgs underlain by a fine to medium silty sand encounteredfrom 10 to 17 ft bgs. At most locations within H1K, a larger amount of silt and clay was encountered

between 20 and 24 ft bgs. Below 24 ft bgs, stiff, dense, very low permeability clay was encountered to adepth of 54 ft bgs. The surficial aquifer at the site is located approximately 7 to 24 ft bgs, and tends toflow to the southeast.

In 2000 and 2001, an Interim Remedial Action using chemical oxidation was conducted inthe source area (CH2MHill and J.A. Jones, 2002). Groundwater sampling indicated that dissolved-phaseconcentrations rebounded following each treatment. In March 2002, a site characterization samplingeffort was performed (CH2MHill and J.A. Jones, 2002) to redefine the magnitude of contamination. Themaximum saturated soil concentrations detected were 1,1,1-TCA at 25,300 g/kg; PCE at 4,360 g/kg;

19


30/54

Figure 2-16. Site H1K, NAS Jacksonville

20


31/54

and TCE at 60,100 g/kg. The maximum groundwater concentrations from a monitoring well weredetected in MW-8 [PCE at 173 g/L, TCE at 5,520 g/L, and cis-1,2-DCE at 1,350 g/L). The highestand second highest total VOC concentrations were detected in injection wells IW-1D (82,340 g/L) andIW-6D (45,782 g/L), respectively. These contaminant concentrations indicate the potential presence ofDNAPL. The baseline TCE contamination plume is shown on Figure 2-17.

The horizontal extent of contamination is approximately 1,450 ft2 with a thickness of 18 ft(saturated zone), resulting in a total volume of 967 yd3 of soil. The estimated mass ranges between 42and 125 lb with the statistical average mass centered at 61 lb.

2.3.2 Technology Implementation. Bench-scale treatability testing indicated that the applicationof NZVI could degrade chlorinated organics present at the H1K site with a removal efficiency between 96and 98% with an applied iron concentration of 1.25 to 13.75 g/L. Test results indicated that generation ofundesirable daughter products (i.e., DCE and VC) from the reduction process was insignificant (TtNUS,2003a). NZVI particles used for this demonstration were manufactured by PARS Environmental Inc. andwere 50 to 300 nanometers (109 meter) in diameter. The particles are BNP and consisted of 99.9% ironand 0.1% palladium and polymer by weight.

NZVI was emplaced using two mechanisms: (1) strategic direct-injection into known hot spotsusing direct-push technology (DPT), and (2) a closed-loop recirculation process (Tetra Tech, Inc.,2005). Direct injection of the nanoscale iron using DPT was employed first at 10 hot spot locations. Arecirculation system was used to distribute the NZVI in the rest of the suspected source zone.

The design of the recirculation system consisted of four injection and three extraction wells,including two existing injection wells for the initial NZVI injection (IW-1 and IW-6). Because theviscosity of the NZVI suspension is similar to groundwater (due to the low iron concentration) the waterwas introduced into the aquifer via gravity flow only. Injection pipes had drilled slots to allow dischargeof the iron into targeted depth intervals that were characterized to have elevated contaminantconcentrations.

For injection via DPT, the iron suspension was diluted to 10 g/L and injected directly into theDPT boreholes using pumps from 7.5 to 23.0 ft bgs, equating to approximately 4.2 lb of iron injected ineach borehole.

During the first recirculation event, water was recirculated into the four injection wells(H10MW-28, -29, -30, and -31) and the two existing chemical oxidation injection wells (IW-1S and IW-6D) via gravity. The recirculation system was operated continuously for approximately 23 hours usingthese wells. Based on bench-scale treatability study results (TtNUS, 2003b), the applied iron concentra-tion was initiated at 2 g/L and later increased to 4.5 g/L based upon field observations indicating the ironwas being accepted by the aquifer without clogging or backing up in the wells. The chemical oxidationinjection wells (IW-1S and IW-6D) were used for injection for a total of 17 hours until they no longeraccepted material.

During the second recirculation event, water was only recirculated into the four new injectionwells (H10MW-28, -29, -30, and -31) for approximately 21.5 hours. The applied iron concentrationremained at 4.5 g/L. The chemical oxidation injection wells (IW-1S and IW-6D) were not used forinjection due to sediment buildup encountered during the first recirculation event.

21


32/54

Figure 2-17. Baseline Contaminant Levels and TCE Contours

22


33/54

2.3.3 Performance Evaluation Approach. Short-term performance monitoring was conductedwith groundwater samples collected within 6 weeks after NZVI injection from a select number of wells.Chemical analysis included Target Compound List VOCs, geochemical parameters, and other parameters.These parameters were analyzed both in the field and in an off-site laboratory.

Longer-term performance monitoring was conducted between 2 months and 1 year after

injection. This phase of monitoring evaluated the longer-term performance of the remedial system in thesource area and within the dissolved-phase plume. The remedial goal established for the study was toreduce the total site contaminant mass by 40 to 50%. In addition to the groundwater sampling discussedin the previous sections, soil and groundwater sampling was conducted in the fall of 2004 to assist inlonger-term evaluation of the treatability study (TtNUS, 2003c).

2.3.4 Technology Performance (Results). Results of samples collected 22 weeks after injectionindicated that the iron recirculation process fostered favorable mass transfer from the sorbed and potentialimmiscible phases into the dissolved-phase. This increase was followed by rapid reductions ranging from65 to 99% of concentrations of parent VOCs in many wells within 5 weeks.

CVOC concentrations for source zone wells H10MW37, H10MW34, and H10MW32 during

the monitoring period are shown on Figures 2-18, 2-19, and 2-20, respectively. Daughter products (pri-marily cis-1,2-DCE) of the parent VOCs (primarily TCE) were detected in all of the sampled wells. Insome source zone wells (Figures 2-18 and 2-19) these daughter product concentrations increased sharplyand subsequently decreased. This was followed by a rise in innocuous byproducts (e.g., ethene andethane). In several source zone wells (Figures 2-18 to 2-20), cis-1,2-DCE levels increased substantiallyover time. This indicates that the iron may not have distributed well in all parts of the source zone or notenough iron mass may have been injected in the treatment zone. The predominance of anaerobicreductive dechlorination products (e.g., cis-1,2-DCE) indicates that much of the reduction in dissolvedTCE levels may have occurred through microbial action or hydrogenolysis, rather than through abioticreduction (beta-elimination).

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

Nov-03

Jan-04

Mar-04

Apr-04

Jun-04

Aug-04

Sep-04

Nov-04

Dec-04

Feb-05

Concentration(g/L)

TCE

cis-1,2-DCE

Vinyl Chloride

Figure 2-18. TCE,cis-1,2-DCE, and Vinyl Chloride Concentrations in Source Area

Monitoring Well H10MW37

23


34/54

0

2,000

4,000

6,000

8,000

10,000

12,000

Nov-03

Jan-04

M

ar-04

A

pr-04

Jun-04

Aug-04

Sep-04

Nov-04

Dec-04

Feb-05

Concentration(g/L)

TCE

cis-1,2-DCE

Vinyl Chloride

Figure 2-19. TCE,cis-1,2-DCE, and Vinyl Chloride Concentrations in Source AreaMonitoring Well H10MW34

0

10,000

20,000

30,000

40,000

50,000

60,000

Nov-03

Jan-04

Mar-04

Apr-04

Jun-04

Aug-04

Sep-04

Nov-04

Dec-04

Feb-05

Concentration

(g/L)

TCE

cis-1,2-DCE

Vinyl Chloride

Figure 2-20. TCE,cis-1,2-DCE, and Vinyl Chloride Concentrations in Source AreaMonitoring Well H10MW32

24


35/54

Detections of compounds such as ethane/ethane and acetylene/C4-hydrocarbons provide someindication that some abiotic degradation via beta-elimination was a secondary treatment pathway.TCE concentrations in a well (H10MW39) located approximately 20 ft downgradient of the targettreatment zone (source zone) were reduced up to 99%. This indicates that some of the injected NZVIcould have migrated outside the treatment zone through preferential pathways.

Figure 2-21 shows that ORP declined to approximately

200 mV immediately after NZVIinjection. The resulting reducing conditions may have been strong enough to stimulated anaerobicbiodegradation and hydrogenolysis, but may not have been strong enough to cause substantial abioticreduction (beta-elimination). Within about 12 weeks, ORP levels rebounded considerably, indicating thatthe NZVI was dissipating. However, the groundwater remained anaerobic for over a year following theiron injection, indicating that conditions suitable for biodegradation continued for a substantially longtime. Groundwater pH levels remained relatively unchanged throughout the demonstration (Figure 2-22),indicating that the NZVI may not have induced strongly reducing conditions suitable for abiotic reduc-tion. Although microbial activity can result in generation of CO2 and a concomitant suppression of pH,there are not enough oxidized species (DO, nitrate, etc.) in the native groundwater for this effect to besignificant (see Table 2-2). Therefore, some increase in pH would be expected following iron injection.

-350

-280

-210

-140

-70

0

70

140

Nov-03

Jan-04

Mar-04

Apr-04

Jun-04

Aug-04

Sep-04

Nov-04

Dec-04

Feb-05

ORP,mV

H10MW32

H10MW34

H10MW37

Figure 2-21. ORP Observed in Monitoring Wells H10MW32, H10MW34, and H10MW37

Tables 2-2 and 2-3 show the trends in several key groundwater parameters measured before(baseline) and after (performance sampling) NZVI injection, as well as approximately one year afterwards(long-term). The performance sampling values in these tables are the extreme values of these param-eters that illustrate the maximum treatment achieved in the aquifer. The wells are grouped into treatmentzone wells (Table 2-2) and extraction wells (Table 2-3); the objective is to see how well the NZVI wasdistributed to wells other than the extraction wells, towards which the NZVI was drawn by a forcedgradient. Interestingly, although dissolved iron (NZVI) levels clearly were higher in the extraction wells

25


36/54

0

1

2

3

4

5

6

7

8

9

10

H10M

W32

H10M

W34

H10M

W37

pH

12/29/03

1/14/04

1/23/04

1/28/04

2/4/04

2/17/04

3/1/04

3/25/04

6/24/04

9/22/04

12/14/04

Figure 2-22. pH Observed in Monitoring Wells H10MW32, H10MW34, and H10MW37

Table 2-2. Key Groundwater Parameters in Treatment Zone Wells

Treatment Zone Wells

H10MW08 H10MW37Parameter

Baseline

Post-TreatmentPerformance

Sampling

Long-TermMonitoring

(Dec04) Baseline

Post-TreatmentPerformance

SamplingMonitoring

(Dec04)

DO (mg/L) 0.08 NS 1.53 1 0.05 0.5

ORP (mV) 221.5 NS 231.7 3.2 226.1 153.2

pH 6.07 NS 6.01 5.81 6.1 6.18

TCE (1/L) 27 NS 210 5,100


37/54

Table 2-3. Key Groundwater Parameters in Extraction Wells

Extraction Wells

H10MW32 H10MW33 H10MW34

Parameter

Baselin

e

Post-Injec

tion

Performa

nce

Samplin

g

Post-Treatment

Monitoring

(Dec04

)

Baselin

e

Post-Injec

tion

Performa

nce

Samplin

g

Post-Treatment

Monitoring

(Dec04

)

Baselin

e

Post-Injec

tion

Performa

nce

Samplin

g

Post-Treatment

Monitoring

(Dec04

)

DO (mg/L) 0.2 0.35 1.9 0.53 NS 2.41 0.42 0.05 0.3

ORP (mV) 3.4 254.2 123.2 146.2 NS -161.8 111.4 256.8 152.8

pH 5.96 6.03 6.02 5.86 NS 6.61 5.77 6.26 6.43

TCE (g/L) 26,000 17,000 18,000 6 NS 1.4 1,100 36


38/54

2.3.6 Regulatory Issues. This demonstration project was conducted under the oversight of regula-tory personnel from the Florida Department of Environmental Protection (FDEP) and the United StatesEnvironmental Protection Agency (U.S. EPA) but with minimal regulatory permit requirements andconstraints.

2.3.7 Discussion. The demonstration showed that NZVI injection is a promising technology for

source zone treatment. The following conclusions and lessons learned can be obtained:

NZVI injection caused a substantial reduction in TCE levels in several sourcezone wells.

ORP reduction was experienced in most monitoring wells in the source zone,indicating that the direct-push and recirculation methods of injection workedrelatively well. Some migration of NZVI outside the treatment zone may haveoccurred through preferential pathways.

Performance data indicate that abiotic reduction of CVOCs (via beta-elimination)was a minor factor in the decline of contaminant concentrations.

The substantial increases in cis-1,2-DCE and 1,1-DCA indicate that microbially-driven anaerobic reductive dechlorination and hydrogenolysis may have played aprimary role in the CVOC treatment.

The NZVI injected did not create the strongly reducing conditions (ORP of400 mV or lower) necessary to generate substantial abiotic degradation of TCE.

One possibility is that the NZVI was passivated before injection when it wasmixed with oxygenated water (groundwater extracted from one of the wells wasused to prepare the iron slurry). NZVI has a very small particle size, is highlyreactive, and can react rapidly with oxygenated species (e.g., DO, nitrate, etc.) inmost water supplies.

Another possibility is that insufficient iron may have been injected. Iron massneeds to be determined based on iron-to-groundwater (or iron-to-soil) ratio,rather than iron-to-contaminant ratio (an ORP of < 400 mV must be achieved inthe target treatment volume). To some extent, NZVI migration outside the treat-ment zone also reduced the mass of iron in the targeted zone.

2.4 Naval Air Engineering Station Lakehurst

2.4.1 Site Description. NAES Lakehurst is located in Jackson and Manchester Townships, OceanCounty, New Jersey, 14 miles inland from the Atlantic Ocean. NAES Lakehurst covers 7,382 acres andis within the Pinelands National Reserve, the most extensive undeveloped land tract of the Middle

Atlantic Seaboard. The site location is shown on Figure 2-23 (Environmental Chemical Corp., 2004).

This project involves two areas with the highest groundwater contaminant concentrationswithin the northern plume (Figure 2-24) and the southern plume (Figure 2-25) in NAES Lakehurst,Areas I and J. The principal contaminants found in the groundwater at Areas I and J include PCE, TCE,1,1,1-TCA, and degradation products such as cis-DCE and VC. The contamination extends vertically70 ft below the groundwater table. The largest amount of contamination is located in the zone from 45 to60 ft below the groundwater table.

28


39/54

Figure 2-23. NAES Lakehurst Site Location Map

29


40/54

Figure 2-24. Total VOC Concentrations in Monitoring Wells Baseline (Northern Plume)

30


41/54

Figure 2-25. Total VOC Concentrations in Monitoring Wells Baseline (Southern Plume)

31

nanoscale zerovalent iron for source remediation

Documents